#325 - 📑 Journal Club - The Complete Episode from June 29th 2025

Hello friends 👋

In this week’s Journal Club, Ben and Daphna unpack a series of recent studies exploring outcomes in neonatology—from long-term mortality after severe neonatal morbidity to short-term feeding strategies in preterm infants.

They open with a large population-based Swedish study showing that infants who experience severe neonatal morbidities face elevated mortality risks well into adolescence—especially those with neurological complications. The discussion highlights how early-life diagnoses carry weight far beyond the NICU, and how long-term support systems may not be fully equipped to manage that risk.

Next, the hosts examine a small German crossover trial on prone positioning, revealing that even simple changes in posture may cut hypoxemic episodes in half. They follow this with a randomized trial comparing dopamine and norepinephrine for neonatal septic shock, a data-heavy look at the nuanced physiology behind first-line interventions.

The episode rounds out with studies on cold milk for feeding dysphagia, late-onset sepsis risk with hydrocortisone, the impact of tele-neonatology on cooling time in HIE, and a new meta-analysis on kangaroo care’s role in reducing infection.

It’s a pragmatic, fast-paced overview of recent literature shaping how we care for vulnerable newborns.

----

The articles covered on today’s episode of the podcast can be found here 👇

Severe Neonatal Morbidity and All-Cause and Cause-Specific Mortality Through Infancy and Late Adolescence.

Graham H, Johansson K, Persson M, Norman M, Razaz N.JAMA Pediatr. 2025 Jun 10:e251873. doi: 10.1001/jamapediatrics.2025.1873. Online ahead of print.PMID: 40493844

Prone Positioning Was Associated With Less Hypoxemic Events and Improved Feeding Tolerance in Preterm Infants.

Bohnhorst B, Lutz E, Pirr S, Peter C, Böhne C.Acta Paediatr. 2025 May 26. doi: 10.1111/apa.70153. Online ahead of print.PMID: 40418109

Prophylactic hydrocortisone and the risk of sepsis in neonates born extremely preterm.

Baud O, Lehert P; PREMILOC study group.Eur J Pediatr. 2025 Jun 14;184(7):419. doi: 10.1007/s00431-025-06248-9.PMID: 40515786 Clinical Trial.

From Warm to Cold: Feeding Cold Milk to Preterm Infants with Uncoordinated Oral Feeding Patterns.

Ferrara-Gonzalez L, Kamity R, Htun Z, Dumpa V, Islam S, Hanna N.Nutrients. 2025 Apr 26;17(9):1457. doi: 10.3390/nu17091457.PMID: 40362766 Free PMC article. Clinical Trial.

Impact of teleneonatology on time to goal temperature in outborn neonates with hypoxic-ischemic encephalopathy requiring therapeutic hypothermia.

Kaczor M, Hentz R, Youssef PE, Fine A, Fang J.J Perinatol. 2025 May 29. doi: 10.1038/s41372-025-02324-y. Online ahead of print.PMID: 40442292

Norepinephrine versus Dopamine for Septic Shock in Neonates: A Randomized Controlled Trial.

Mazhari MYA, Priyadarshi M, Singh P, Chaurasia S, Basu S.J Pediatr. 2025 Jul;282:114599. doi: 10.1016/j.jpeds.2025.114599. Epub 2025 Apr 17.PMID: 40252959 Clinical Trial.

All-cause mortality and infection-related outcomes of hospital-initiated kangaroo care versus conventional neonatal care for low-birthweight infants: a systematic review and meta-analysis.

Minotti C, Jost K, Aghlmandi S, Schlaeppi C, Sieswerda E, van Werkhoven CH, Schulzke SM, Bielicki JA.Lancet Child Adolesc Health. 2025 Jul;9(7):470-483. doi: 10.1016/S2352-4642(25)00130-0. Epub 2025 May 26.PMID: 40441171 Free article.

----

Watch this week's Journal Club on YouTube 👇

----

The transcript of today's episode can be found below 👇

Ben Courchia MD Hello everybody, welcome back to the Incubator Podcast. We're back today for an episode of Journal Club. Daphna, how are you?

Daphna Yasova Barbeau MD I'm doing well. I have some articles left over from last time because you had such big hitters, discussing the PDA and whatnot.

Ben Courchia MD Yeah, that article definitely made the rounds. There's been a lot of discussion about it. But like we said last time, it's nice to have these goalposts saying, "This is where we're going to plant the flag for now," and then let's see if we can move it. It's been a busy time for us, especially from the editorial standpoint. It’s a busy time on the article front when such major reports come out. As you’ve seen on the podcast, we like to review smaller studies sometimes, especially if they ask interesting questions or have very unique findings. Unfortunately, these often take a backseat when something like the AAP report on PDA comes out. So, our list has been piling up, and there are definitely a lot of articles to review. I'm very excited about today—lots of interesting topics to touch on. Without much further ado, let's get started.

I just wanted to give two quick updates to the audience. We're processing the submissions for the annual giveaway. If you haven't heard from us, it's not that we won’t respond—we’re just tallying all the entries. We are, notoriously, an inefficient group.

Daphna Yasova Barbeau MD Well, I think we prioritize Journal Club, board reviews, and interviews—but we'll get there. Yeah, that's right.

Ben Courchia MD Yeah. We'll get there, but I'm saying it’s our fault. It's not anyone else's. Also, the Delphi Conference is open for registration, and we’ll be sharing more content about the speakers and the lineup. It's going to be very exciting. You're not going to want to miss this. We're keeping it an intimate type of conference, not unlimited in size. It's limited seating. So, if you're planning to attend January 26–28, 2026, you'll be coming to sunny Florida in the middle of winter. It'll be pleasant. We’ve never invested as much time and effort into the lineup and curation of the content. It will be exceptional.

Daphna Yasova Barbeau MD People are already registering, so that’s exciting.

Ben Courchia MD Oh yeah, and that’s why I wanted to remind people. I don’t want someone to say later, “I wanted to register, but it was too late.” There’s really nothing we can do. We pick venues that have a certain number of seats. We're not being elitist about it—it’s just a fixed number. After that, we can’t accommodate more.

Daphna Yasova Barbeau MD Yeah, and I think it's good for people to know that as they're planning. We’ve selected a hotel—the Flow Hotel—which is very South Florida, but also very comfortable. There are options for people who want to share rooms or travel together. We want everyone to be really comfortable and have flexibility—whether you come with a group or bring your family, we can accommodate that.

Ben Courchia MD Just so people know, the rooms are very serene—lots of wood, very relaxing, nice views. They're almost like little apartments, so you have some independence. There’s a great gym, a beautiful pool, and even an outdoor yoga area. When we say it’s going to be a relaxing conference where you leave with more energy than you came in with, we really mean that. We pay a lot of attention to that.

All right, let’s get going with Journal Club. The first paper I wanted to review is a very interesting one from JAMA Pediatrics. It's called "Severe Neonatal Morbidity and All-Cause and Cause-Specific Mortality Through Infancy and Late Adolescence." The first author is Hillary Graham. It's a study out of Sweden examining the long-term mortality implications of severe neonatal morbidity. This is particularly relevant as we see more infants surviving the neonatal period, many of whom face significant health challenges. While survival rates have improved dramatically, we're seeing more infants with morbidities that have lasting consequences. The authors point out that identifying key morbidities during the perinatal and neonatal period could help us target those at the greatest risk for early-life mortality.

It’s terrifying to talk about early-life mortality, but it’s something we can’t shy away from. In the paper, they refer to this concept of SNM—Severe Neonatal Morbidity—during the first 27 postnatal days. It’s an important health indicator, but it includes a variety of diagnoses and is often inconsistently defined across studies. Common SNMs include IVH, extremely and very preterm birth, sepsis—all associated with increased risks of morbidity and mortality.

Previous research has shown striking findings. One study demonstrated that infants with certain neonatal morbidities have a 10-fold higher risk of death in their first year. Another found that SNM is a strong risk factor for infant and child mortality up to five years of age and is significantly associated with higher mortality from conditions affecting the endocrine, nervous, respiratory, and circulatory systems. What’s been lacking is long-term follow-up for SNM subtypes beyond five years of age and any sibling-controlled analyses. This Swedish group decided to examine the association between SNM during the neonatal period and all-cause and cause-specific mortality from infancy to adolescence using both population-based and sibling-controlled analyses.

They did a population-based study from 2002 to 2021 in Sweden, leveraging their comprehensive registry system by cross-linking the medical birth register, the Swedish Neonatal Quality Register, the National Patient Register, and the Cause of Death Register using national ID numbers. These registries are incredibly robust—the National Patient Register includes diagnostic codes since 1987 and outpatient care since 2001. The Cause of Death Register captures about 99% of all deaths in Sweden since 1952. Quite incredible.

The final study population was about 2 million singleton neonates. The main exposure was the occurrence of SNM within the first 27 days of life. SNM was defined as a composite event consisting of five categories and 19 subtypes:

• Respiratory morbidity: RDS, congenital pneumonia, BPD, other respiratory disorders, pneumothorax requiring drainage

• Infections: Sepsis, bacterial meningitis

• Neurological morbidities: Grade III–IV IVH, HIE, seizures, cerebral infarction, PVL, hydrocephalus

• Respiratory procedures: Resuscitation with intubation, mechanical ventilation

• Other: NEC, intestinal perforation, ROP, birth trauma

They examined all-cause mortality, categorized as:

• Infancy (28 days–11 months)

• Childhood (1–4 years)

• Late childhood (5–9 years)

• Adolescence (10+ years)

Covariates included gestational age, birth weight, sex, SGA status, Apgar score, birth year, and any major malformations. To address potential familial confounders, they also conducted a sibling-controlled analysis using sibling pairs discordant for SNM exposure, adjusting for infant and maternal factors like age, parity, smoking, cohabitation, education, BMI, pre-gestational diabetes, and comorbidities.

Let’s look at the results. Of the 2 million infants, 49,225 (2.4%) had SNM in the first 27 days. Infants with SNM were more likely to be preterm, male, SGA, have lower 5-minute Apgar scores, and have major malformations. SNM prevalence was higher among infants born  to mothers who were to young (<19 years old), less educated, smokers, living alone, or those with higher BMI, pre-gestational diabetes, or hypertension.

During a follow-up of up to 21.2 years, 3,618 children died—an overall mortality rate of 0.16 per 1,000 person-years. Among children with SNM, the mortality rate was 1.81 per 1,000 person-years, compared to 0.13 in those without SNM. In adjusted models, SNM was associated with a nearly six-fold higher hazard ratio for mortality. Some morbidities had even more dramatic effects. Neurological conditions were associated with an 18-fold increase in mortality. There were also significant sex differences; we tend to think that survival rates for females are generally higher than males in the NICU. But females with SNM had a 7.28 hazard ratio, compared to ~5 for males. Among term infants with SNM, the adjusted hazard ratio was 7.16, compared to 3.51 for preterms—suggesting that when term infants develop SNM, consequences can be even more dramatic. The effect of SNM tapered over time but persisted: 7.66 at 28 days–11 months, 5.2 from 1–4 years, 4.98 from 5–9 years, and 2.28 after 10 years. Even at 10+ years, infants with neurological or respiratory procedures continued to have elevated mortality.

The sibling analysis, involving 32,000 children with SNM vs. their unaffected siblings, found a similar adjusted hazard ratio of 4.33, confirming much of the population analysis. However, respiratory conditions like RDS and BPD showed attenuation in the sibling analysis—possibly suggesting a genetic or familial component. For example, a family with a history of severe asthma might have a baby at higher risk of long-term complications after NICU morbidity than one with healthy lungs.

In conclusion, the authors found that infants with SNM who survive the neonatal period experience increased mortality risks into childhood and also adolescence. Neurological morbidities and respiratory procedures are the major contributors to both short-term and long-term mortality. Efforts to further prevent neonatal morbidity, ensure early identification, and long-term follow up are key to reducing this risk and guiding families. Daphna, what do you think? It’s the kind of data I’d be terrified to present to families in the NICU.

Daphna Yasova Barbeau MD Yeah. Well, I mean, I think it's not unexpected information, but I think it's really important when we talk about long-term outcomes. It's kind of this cumulative effect, you know? It's not just gestational age. It's not just birth weight. And I sometimes feel like we have this sense, especially for people who aren't doing follow-up, like, wow—we got this very sick baby to the endpoint. They're going home. And that's great. That's wonderful. I mean, we are making extraordinary strides in getting babies to discharge. But that cumulative effect of these morbidities and the impact on long-term mortality is impressive. I think we need to keep that in mind, and we have to keep working on how we safely transition these babies home and what types of resources they have. I mean, you and I both know in our area, for example, these medically complex kids—home nursing is a disaster. So we really have to figure out how to optimize those things. I was especially, like you, interested in the sibling data. You know, we've got a special event for Delphi specifically for multiples, and so it was interesting to see that as well.

Ben Courchia MD I think that, like you said, when you compound everything—Sweden has a very robust social health care system. So when you consider how difficult it is, at least speaking for ourselves, for parents to actually attend multiple appointments and get access to the different subspecialists—with insurance issues, lapses, and so on—it definitely probably means that odds are higher in a setting like ours. Again, this is Sweden, one of the leading countries in neonatal care. We're not talking about an underserved or emerging field of neonatology. These are the experts. These are the Scandinavian teams that we strive to emulate.

Daphna Yasova Barbeau MD Yeah, and the social networks there are quite robust with parental leave and universal health care.

Ben Courchia MD I wouldn’t even know how we would capture that kind of data here in the U.S. Would we even be able to? How many times do you hear about a baby who has somehow appeared at an outlying ER that may not be affiliated with your healthcare system? It’s like—oh my God. So anyway, very interesting data. Where are you taking us next?

Daphna Yasova Barbeau MD All right. Well, I have this paper that, hopefully, can help reduce some of the morbidities. It’s from Acta Paediatrica and is entitled “Prone Positioning Was Associated With Fewer Hypoxemic Events and Improved Feeding Tolerance in Preterm Infants.” Lead author is Bettina Bonhorst, and senior author is Carolyn Bone. This is coming out of Germany. Basically, what they wanted to look at was how positioning affects cardiorespiratory events—apnea, bradycardia, and hypoxemia—which we know are very common in preterm infants. We’re learning that the frequency and especially the severity of hypoxemic events are impacting neurodevelopmental outcomes.

It’s a small study—I’ll start by saying that—but between May 2016 and July 2018, they conducted a single-center, prospective, randomized, two-arm crossover study at Hanover Medical School, their tertiary NICU. They enrolled preterm infants with a gestational age of less than 32 weeks at birth and less than 34 weeks postmenstrual age at the time of the study. The babies had to be experiencing more than six episodes of hypoxemia (sats <80%) and bradycardia (<80 bpm) within six hours. That’s quite a lot. They excluded babies with congenital malformations, intestinal pathology, or those needing noninvasive positive pressure ventilation (like NIMV or mechanical ventilation). So all of these babies were extubated. In terms of breathing support, they were either on room air, receiving supplemental oxygen, or on high-flow nasal cannula or CPAP. Importantly, the ventilatory support did not change for any baby over the study period, although some may have needed intermittent adjustment during a prolonged event.

Patients were randomly assigned to either Group A (prone) or Group B (supine) as their starting position. Each patient spent 12 hours in the starting position while events were recorded, followed by a 12-hour washout period, and then switched positions.

They had 48 patients. Median gestational age was 26.9 weeks (range 24.4 to 30.3), and mean birth weight was 884 grams ± 250 grams. At the time of study, mean postmenstrual age was 31.7 weeks, and current weight was 1,273 grams. Each group had 24 patients. Demographics were pretty well matched for gestational age. However, the median weight at study entry was significantly lower in Group B (supine): 1,163 grams versus 1,378 grams in Group A. That’s important because Group A is prone, Group B is supine—but the prone group had higher median birth weight.

The total number of hypoxemia and bradycardia events didn’t appear affected by starting position, and there were no carryover effects in secondary outcomes.

The primary outcome was cumulative frequency of hypoxemia and bradycardias per hour. This was significantly decreased—almost halved in the prone position: 17.6 in supine versus 8.3 in prone. This decrease was primarily due to fewer hypoxemic events. The prone position helped both moderate hypoxemia (sats 85–80%) and severe (<80%). The occurrence of bradycardias didn’t differ significantly between the positions.

Secondary endpoints included total time spent in hypoxemia, which was also significantly decreased—37.5 minutes in supine vs. 16.3 minutes in prone. For severe hypoxemia, supine group spent 22.9 minutes vs. 9.1 in prone. There was a trend toward worsening severity in the supine group, but it wasn’t statistically significant. Median duration per event wasn’t statistically different. Baseline parameters like heart rate, respiratory rate, periodic breathing, and supplemental O2 need didn’t differ significantly. However, median and minimum baseline SpO2 were significantly increased in the prone group: median saturation was 96.3% in supine vs. 97.9% in prone. Minimum saturation was 92.9% in supine vs. 95.1% in prone.

They also looked at gastric residuals. We’re moving away from checking these, but the findings were interesting. Residuals were significantly lower in the prone position: median was 3.75% of the previous meal in supine, versus nothing in prone. Maximum residuals were also lower—12% in supine vs. 6% in prone.

However, the frequency of apneas increased in the prone position: 1.2 per hour in prone vs. 0.4 in supine. That was statistically significant, although the duration per event didn’t differ.

So prone positioning offered many benefits, but this was in slightly bigger babies—which we know impacts NICU outcomes. I think it’s interesting. Some units still don’t use prone positioning at all. I found this a useful paper. Thoughts?

Ben Courchia MD Yeah, I mean, it’s always interesting when a small study like this reaches statistical significance. It means the difference they found was staggering. I think this is something to look at. As you mentioned, it's a small sample, and that’s okay. But a 50% reduction in events is not trivial. This data supports what we know about prone positioning. It doesn’t change the fundamental challenge, which is that we can’t leave these babies prone 24 hours a day. But it’s an intervention at our disposal—use it judiciously.

Daphna Yasova Barbeau MD Yeah. I thought the gastric residuals were interesting. Could that help with some of our reflux-type problems? I don’t know—they didn’t prove that, but it’s interesting. We don’t routinely check residuals, but still.

Ben Courchia MD Yeah, very interesting indeed. Okay. The next paper I have for you is “The Risk of Sepsis in Neonates Born Extremely Preterm.” First author is Olivier Beaux, on behalf of the Premiloc Study Group. I had the pleasure of interviewing Olivier with Gabriel Altit on the French edition of the podcast. If you're French-speaking, go listen to it. He’s a brilliant mind who has done a lot of work with Kristi Watterberg on the use of hydrocortisone. Kristi mentioned him several times during our interview with her on the English edition.

What the team at the Premiloc Study Group wants to examine is a critical safety concern around the use of prophylactic hydrocortisone in extremely preterm infants. This is a post hoc analysis of the Premiloc trial that addresses one of the main barriers to implementing prophylactic hydrocortisone as standard of care.

Despite the evidence, hydrocortisone hasn’t been widely adopted, largely because of concerns over long-term outcomes. The increased susceptibility to infection in preterm infants is due to a deficient immune system, immature epithelial barrier, and increased use of invasive devices. The extremely low gestational age neonate (ELGAN) population is particularly at high risk for late-onset sepsis—about 20% experience at least one culture-proven episode prior to discharge.

Prophylactic low-dose hydrocortisone has been shown to improve survival without BPD, reduce need for PDA treatment, and reduce pre-discharge mortality. As a result, it's being introduced in some NICUs. Olivier and his colleagues strongly support this for very vulnerable infants. They also acknowledge that some systematic reviews show the greatest benefit in infants born after 25 weeks but does raise safety concerns. Although some centers have adopted the Premiloc regimen, it’s not standard of care. One concern comes from an individual patient data meta-analysis of main trials, showing increased risk of late-onset sepsis with an odds ratio of 1.34—particularly in infants born at 24–25 weeks, where the odds ratio is 1.87. Surprisingly, this increase didn’t correspond to increased mortality or neurodevelopmental disability at discharge, or at 2 or 5 years. Real-world data have come from retrospective cohorts in the UK, Canada, the US, and more recently Sweden.

So the purpose of this study is to reanalyze Premiloc data to predict late-onset sepsis incidence from baseline conditions, examine perinatal factors, and assess the effect of prophylactic hydrocortisone exposure. This was a post hoc analysis of the original Premiloc trial—a double-masked, placebo-controlled, multicenter RCT. It enrolled extremely preterm infants (<28 weeks) from 21 perinatal centers in France. Infants received either prophylactic low-dose hydrocortisone or placebo within 24 hours of birth for 10 days. The original trial’s primary endpoint was survival without BPD at 36 weeks PMA. The primary outcome of this post-hoc analysis was late-onset sepsis, defined by a positive blood culture or pneumonia with clinical deterioration after day 3, per international definitions. Pneumonia specifically was defined as a chest X-ray showing pulmonary opacities associated with the presence of a positive culture in the airway and clinical deterioration after day of life three. Late-onset sepsis occurring after day seven or day 10—that is, at the end of the hydrocortisone exposure—was considered as an endpoint for the sensitivity analysis only.

Now, the researchers categorized respiratory support at baseline into three categories. Respiratory support at baseline (RSB) was considered mild if the infant was on non-invasive ventilation with less than 30% FiO2, moderate if on mechanical ventilation with FiO2 less than 30%, or severe if on mechanical ventilation with a need for 30% or greater. Secondary neonatal outcomes specifically included the total number of days the infant was mechanically ventilated, age at weaning from respiratory support, PDA, and treatment for PDA (medical or surgical).

In terms of the data, they were able to get most of the babies from the Premiloc cohort. Out of the 523, they were able to analyze 519 of them based on some consents that were pulled, but nothing significant from that standpoint. A relative difference of more than 15% between the compared groups was found for general anesthesia and gestational hypertension only, and there were no other significant differences between other variables in the baseline characteristics between the babies who did receive hydrocortisone versus those who did not.

So let’s get into some of the outcomes. Overall, the proportion of late-onset sepsis reported in the population was 27.2% after day three, 23% after day seven, and 17% after day 10, with a significantly higher proportion in the hydrocortisone treatment group. The case fatality rate likely attributable to late-onset sepsis was 23% in the placebo group and 26% in the hydrocortisone group.

Now, using a stepwise strategy to study the occurrence of late-onset sepsis after day three, the main analysis provided three main predictors at baseline for increased incidence of late-onset sepsis. A decreased risk was observed with increased gestational age, with a relative risk of 0.772 per week. Vaginal delivery was also associated with decreased risk, with a relative risk of 0.611, while perinatal asphyxia showed a trend toward increased risk with a relative risk of 1.66.

Among covariates occurring during the first postnatal week, supplemental steroids given after 10 days of treatment with prophylactic hydrocortisone—but before any form of sepsis occurred—were found to be associated with a decreased risk of LOS. So, the supplemental use of hydrocortisone somehow lowered the risk; the kids who did get supplemental hydrocortisone had a lower risk of developing sepsis.

For the effect of hydrocortisone treatment on the risk of late-onset sepsis, the model predicting the occurrence of sepsis found in the previous section—those three variables—was used. The hydrocortisone effect, adjusting for gestational age, vaginal delivery, perinatal asphyxia, and even supplementary corticosteroids, showed the proportion of late-onset sepsis was 24.5%, with a standard deviation of 9.8% across centers, giving a variation coefficient of 0.33.

The effect of the three main adjustment variables did not change, and adjusted for these covariates, the researchers found a non-significant hydrocortisone effect with a relative risk of 1.04. When they investigated late-onset sepsis occurring between day three and the end of the course (day 10), and adjusted the same model, they found a treatment effect with a relative risk of 0.96—so, almost no difference.

The researchers also performed a competing survival analysis using the Fine and Gray model, considering death and late-onset sepsis as competing endpoints. After adjustment for baseline variables associated with risk of late-onset sepsis—again, gestational age, delivery mode, and perinatal asphyxia—the effect of prophylactic hydrocortisone on late-onset sepsis was found not to be statistically significant, while competing death was reduced by treatment with a hazard ratio of 0.427.

The conclusion of this detailed data analysis is that, considering baseline risk factors, the risk of LOS (late-onset sepsis) is not significantly different between the group that received prophylactic hydrocortisone and the placebo group. The competing survival analysis by Fine and Gray suggests that death was reduced by prophylactic use of hydrocortisone without a significant effect of the treatment on the risk of late-onset sepsis. These results, they say, should be confirmed in larger, real-world, population-based cohort studies.

To me, it’s very interesting because, again, it’s Olivier really standing behind the data and the Premiloc trial. I don’t know how I take this data at face value, but I think the underlying message is the right one, in my opinion. Every baby is not the same. There are baseline risk factors that will make a baby more likely to benefit from this treatment versus others—whether it is gestational age (which we kind of know), vaginal delivery, perinatal asphyxia, and so on.

Those are very interesting to investigate. I think maybe it is a way for the Premiloc team to move away from a wholesale prophylactic use of hydrocortisone and instead say, maybe there is a subpopulation that would truly benefit the most—all the while not being exposed to some of the potential adverse effects that we’ve been concerned about. I think this is very interesting.

Daphna Yasova Barbeau MD Well, I think it’s always complicated when you're interrogating data that wasn’t intended for that purpose. But I mean, when you have studies with such large datasets, I almost feel like you're obligated to do some of this additional work. Everybody’s done the work, the patients have been enrolled, the information is there, and we should be looking at it, if for nothing else than to generate new hypotheses and go from there.

But I do think this is relevant. People are worried that if we give steroids very early to a baby who may be at risk for infection, will we compromise the immune system and increase their likelihood of infection? So I do think this targets some concerns people have about early hydrocortisone. I think one of the groups we think are most likely to benefit are those babies with a history of chorioamnionitis. So I think this is especially interesting as well.

Ben Courchia MD Yeah. And I just want to mention—obviously, I invite everybody who hasn’t read the Premiloc trial, which was published in The Lancet, I believe, in 2016? I think what was interesting in the Premiloc study is the fact that in that particular population, when they reported the severe adverse events, they didn’t see any difference in adverse event rates, except for the fact that babies who received hydrocortisone had a lower mortality rate prior to discharge.

When it comes to sepsis, the numbers were fairly the same: about a 31% rate of late-onset sepsis in the hydrocortisone group versus 25% in the placebo group. You said they're interrogating their data, but I feel like also there’s a sense that other meta-analyses are doing that too—especially the individual patient data meta-analyses. So I think this comes as a response to that. But it is tedious work for sure. And what you do with that information, I don’t know. I’m curious to see what the response on Twitter will be to this, because hydrocortisone divides more than anything in our field.

Daphna Yasova Barbeau MD That’s right. For people who want to listen to our interview with Dr. Watterberg, it was Episode 53. She took a chance on us when we were still pretty early, coming to chat with us on the podcast in that first year. So people can check that out. We were starstruck.

Ben Courchia MD That was a big get.

Daphna Yasova Barbeau MD I had another interesting paper, actually in the journal Nutrients. It's entitled "From Warm to Cold: Feeding Cold Milk to Preterm Infants with Uncoordinated Oral Feeding Patterns." The senior author is Nazee Hanna. This was done in New York—a collaboration between neonatologists at NYU and the University of Arkansas.

The background is kind of interesting. They recognized that in adults with dysphagia, cold liquids enhance swallowing by stimulating sensory receptors in the pharyngeal mucosa. They had previously shown that feeding infants with cold liquid reduced dysphagia in preterm infants. But I think a lot of units are still not using cold feeding, so that’s something they wanted to explore.

They selected infants with concerns for disorganized feeding and compared feeding cold milk versus room temperature milk—not even warm milk, just room temp. I’ll give you the punchline early: the predominant results of the study are negative. But I think that's interesting in and of itself because we spend a lot of time warming milk. So I’ll give you some of the details.

They enrolled preterm infants less than 37 weeks gestational age at birth with a postmenstrual age of 34 to 42 weeks. These were babies who were working on PO feeding at the time of enrollment. They had to have a documented uncoordinated feeding pattern using the Infant-Driven Feeding Scale (IDFS), which they routinely used in their unit. Infants were excluded if they had growth restriction, a history of GI pathology like NEC, gastroschisis, malabsorption, diarrhea, or omphalocele. They also excluded babies with NAS due to its potential effect on feeding behavior.

Each participant was exposed to two oral feeding conditions—either room temperature (68–77°F) or cold temperature (39–48°F)—in a randomized crossover design. So babies started with one, then after a washout of several feeds within a 12-hour period, they received the other feeding temperature. That allowed enough time to return to baseline and reduce changes in clinical status. In the room temperature (control) condition, infants got their usual milk, either formula or breast milk, at room temperature. In the cold condition, milk was never brought to room temperature. Feeding was performed by a trained research team member with infants positioned in a sidelying posture using a slow flow nipple.

They looked for signs of cold stress, such as axillary temperature changes. For babies who already had NG tubes in place, they measured gastric contents’ temperature. They did not place NG tubes solely for the study, to avoid unnecessary trauma. They measured gastric aspirate temperature at baseline, and then for the infants fed cold feedings, measured the temp at at 1, 10, and 30 minutes post-feeding to monitor return to baseline. They also monitored for signs of cold stress like tachypnea, desaturation, apnea, bradycardia, increased gastric residuals, and emesis. Additionally, in a subset of infants, they used Doppler ultrasound to evaluate superior mesenteric artery blood flow as a marker of intestinal perfusion, given concerns about feeding cold milk. For the Doppler assessments, they measured one hour before feeding, then 30 and 60 minutes after each feeding.

Feeding performance was evaluated using several parameters assessed at five-minute intervals. They used the Oral Feeding Skills (OFS) score, the Infant-Driven Feeding Scale, and two additional subjective measures: stress cues and need for pacing, using validated stress and pacing scales.

Again, this was a small study—26 preterm infants were included, with five infants undergoing Doppler assessments. The median gestational age at birth was 31 weeks and 6 days. The median day of life at enrollment was 28 days, with a median postmenstrual age of 36 weeks and 1 day. All babies had some prior oral feeding experience, with a median of eight days of partial oral feeding before enrollment. Fourteen infants started with room temperature feeding; twelve started with cold. As expected with a crossover design, all received both.

There was no significant reduction in axillary body temperature. Gastric temperature measurements (only performed in infants who already had NG tubes, so only measured in seven babies) showed that within one minute after cold milk feeding, the temperature increased to an average of 83.3°F. That means it warmed almost immediately in the stomach. The gastric content temperature after cold feeding was only 3.6°F lower than after room temperature feeding at the one-minute time point, which was statistically significant. But in terms of clinical signs of cold stress—no differences were seen between groups. In the small Doppler cohort, there were no significant differences in superior mesenteric artery blood flow between feeding types.

Now in terms of feeding assessments: there was no statistically significant difference in OFS scale levels or other objective measures like volume transferred or rate of transfer. Subjective assessments of stress cues and need for pacing also showed no significant difference.

However, one notable finding was that coordination of sucking, swallowing, and breathing—measured by the Infant Dysphagia Feeding Scale—did show a statistically significant improvement with cold temperature compared to room temperature. The mean IDFS quality score was 3.2 for room temperature and 2.6 for cold feeding—lower score is better. So, better coordination during cold feeding.

Individually, 10 of the 21 infants improved by one to three points, 10 showed no change, and only one infant had a one-point decline with cold milk. So while many feeding parameters didn’t change, they hypothesized that it would impact dysphagia, and they did demonstrate a statistically significant improvement there. It’s a small study, but I think they did a good job evaluating balancing measures and things people worry about. And since many of our babies stay a long time just for feeding readiness, it’s worth considering other feeding strategies to potentially improve outcomes.

Ben Courchia MD I couldn't find the table but this is very interesting. I think you shot our enthusiasm a little bit in the beginning by giving the punchline early! Maybe a lot of nurses listening were thinking, "Oh my God, I won’t have to thaw milk as much—this will save a lot of time!" But it’s an interesting approach. Especially with room temperature milk—what do you think? Based on the data, it’s a small study, but if a baby is stuck, would you try room temperature and see if it makes any difference?

Daphna Yasova Barbeau MD Well, you know, it’s interesting. In our unit, we don’t even use room temperature milk—it’s slightly warmer than that. But yes, especially for a baby who is stuck. We have full-term babies still struggling to feed. And plenty of full-term babies at home are getting cold milk. So yeah, I think in a baby who’s really stuck, maybe there are some who would benefit.

Ben Courchia MD Yeah, because I think that’s another thing I didn’t mention. We’re all probably using similar machines—our milk is probably on the warmer side rather than the colder side. And as a human, you feel like you should feed warm milk to your baby. But I’m just curious whether you’d try a cold bottle.

Daphna Yasova Barbeau MD Yeah, I think if you're really stuck—especially in a baby who is, say, term corrected—I think people are still concerned about cold stress or mesenteric ischemia. I don’t think they had the numbers to really look at that.

Ben Courchia MD I mean, it’s a chilled glass of milk—it’s like 5 to 10 degrees Celsius.

Daphna Yasova Barbeau MD Which, as an adult, is delightful. Nobody wants warm milk once you hit a certain age! So I’d try it, especially in a baby who’s stuck at term corrected.

Ben Courchia MD Yeah, especially when you have these interesting interventions where you don’t really need a ton of data to safely try it—like playing soft music in the NICU. It’s like, how much data do you need to justify a lullaby? I think it's interesting. All right—I’ll keep an eye on the babies you try this on and see how they do.

Daphna Yasova Barbeau MD And you know we’ve all been there, right? At home, the milk is there, the baby’s screaming, and you’re like, “I guess I won’t warm it—I’ll just feed it and see how it goes.” So babies are getting cold milk at home.

Ben Courchia MD They are getting cold milk. All right, we have a few more articles to go through. I have two more articles and I’ll give us a bit of a break. I wanted to talk about something we don’t often cover. I really wanted to talk about a paper I found in the Journal of Perinatology. It's called "Impact of Tele-Neonatology on Time to Goal Temperature in Outborn Neonates with Hypoxic-Ischemic Encephalopathy Requiring Hypothermia." The first author is Mark Caxor, and the last author is Jennifer Fang, who is on our lineup for the Delphi Conference. This is a topic we're very interested in. This group out of the Mayo Clinic in Rochester was examining how telemedicine technology can improve outcomes for one of the most time-sensitive interventions in neonatology: therapeutic hypothermia for HIE.

The background is interesting. HIE remains a leading cause of neonatal morbidity and mortality worldwide. Therapeutic hypothermia is pretty much the only intervention currently available, and it relies on prompt initiation to optimize neuroprotective effects. This can be logistically challenging for neonates born at non-tertiary centers who require transport to regional centers—the "outborn" population. Outborn neonates requiring hypothermia for HIE have been reported to have higher rates of death, more frequent seizures requiring treatment, and more frequent gastrostomy tube placement compared to neonates born at hospitals with NICUs equipped to deliver therapeutic hypothermia. This highlights the critical importance of optimizing the care pathway for these vulnerable infants.

Now, tele-neonatology—real-time audio-video telemedicine consultation with a neonatologist provided to surrounding community hospitals—has the potential to accelerate this process by offering expert assessment and recommendations within minutes after birth. It's been shown to improve the quality of high-risk newborn resuscitation. A previous retrospective cohort study showed that perinatal asphyxia or encephalopathy was the admission diagnosis for 15% of neonates who underwent consultation.

While the use of telemedicine to assess neonatal encephalopathy has been shown to be feasible, it's still unknown whether it could improve the therapeutic hypothermia process or early health outcomes in these neonates.

The objective of this study was to assess whether tele-neonatology is associated with earlier achievement of goal temperature in outborn neonates with HIE who underwent cooling. A secondary objective was to determine whether tele-neonatology was associated with reduced early neurologic morbidity or in-hospital mortality.

This was a retrospective observational study of neonates born at community hospitals and transferred to the Mayo Clinic for therapeutic hypothermia between 2013 and 2022. The Mayo Clinic has a 24-bed Level III NICU and a 34-bed Level IV NICU, with approximately 800 admissions a year. All neonates discussed here underwent cooling at the Level IV NICU. Decisions to initiate therapeutic hypothermia were made using criteria from the NICHD and the Neonatal Research Network trials on cooling. The exposure of interest was whether the consultation occurred via tele-neonatology (a video telemedicine service available 24/7) or by phone. During the study period, subjects were admitted from 19 community hospitals. Twelve participated in Mayo Clinic’s tele-neonatology program for at least part of the study period; the remaining seven did not. At the tele-neonatology sites, referring physicians could choose between phone or video consults. At non-participating sites, consultations were by phone only.

In the phone-only control group, transport teams could be dispatched after identifying the need for a higher level of care—but that required an extra step: the referring physician had to make a separate call to the transfer center. In contrast, tele-neonatology enabled the neonatologist to activate transport during the same consultation call, avoiding additional delays.

The primary outcome was time from birth to achieving a core temperature between 33 and 34°C (the target range for therapeutic hypothermia). Secondary outcomes included a composite of early neurologic morbidity or mortality: seizures, significant brain injury on MRI, or death before acute discharge or transfer.

Now to the results: 77 neonates met eligibility criteria. Forty-two (55%) were in the tele-neonatology group, and 35 (45%) were in the control group. Baseline characteristics were generally comparable between the groups, with the exception of race: 71% of the tele-neonatology group were white, compared to 83% in the control group. Advanced airway placement occurred more frequently in the tele-neonatology group (31% vs. 20%). Both groups had similar rates of chest compressions (26%) and epinephrine use (7% vs. 6%). Surprisingly, the median time from transport team dispatch to arrival at the birth hospital was similar between groups: 50 minutes for the tele-neonatology group versus 55 minutes for the control group. I would have expected a larger difference. The median time from birth to initiation of the telemedicine consult was 13 minutes, and the average duration of the consult was 56 minutes.

For the primary outcome: neonates in the tele-neonatology group achieved goal temperature in a median of 3.4 hours, compared to 4.6 hours in the control group—a mean ratio of 0.75. This remained statistically significant after adjusting for patient acuity using blood gas pH (adjusted mean ratio of 0.76). They also performed a sophisticated causal inference analysis using directed acyclic graphs to understand the mechanisms behind this improvement. Two key mediating factors emerged:

1. Time from birth to activation of the transport team was significantly shorter in the tele-neonatology group—median 23 minutes versus 61 minutes. We're all struggling with the same thing - we activate it, but it takes some time to get there.

2. Initiation of passive cooling before transport team arrival was more frequent in the tele-neonatology group—57% versus just 14% in the control group.

Regarding concerns about overcooling, 7% of infants in the tele-neonatology group and 6% in the control group were below the target temperature upon transport team arrival—essentially no difference. As for secondary outcomes: there was no detectable difference in early neurologic morbidity or mortality—57% in the tele-neonatology group versus 49% in the control group.

The authors concluded that tele-neonatology reduces time to goal temperature in outborn infants receiving therapeutic hypothermia. These findings highlight the potential of telemedicine to improve the timely delivery of care to vulnerable infants.

Now, some might say 3.4 vs. 4.6 hours doesn’t sound like a huge difference—but that’s more than an hour. And while it’s still within the six-hour window, emerging data suggests that earlier initiation—by even one hour—may yield incremental neuroprotective benefits. I was talking with a friend about this recently—maybe tele-neonatology is ahead of the curve. Even if we don’t yet have definitive data showing that cooling at one hour versus four hours is dramatically better, we already have a way to make it happen. If you activate a consult system via telehealth, you might achieve that earlier initiation. I think this is a very interesting paper. One of the reasons we’re highlighting this topic at Delphi is because it’s coming. And if you’re reading this thinking, "If I take a job somewhere, I might be expected to provide tele-neonatology consults to outlying hospitals," you’re right. I don’t know about you, Daphna—I’ve never done this before.

Daphna Yasova Barbeau MD Yeah, I mean, we did this extensively at the University of Florida. And actually, Mike Weiss has been doing tele-Sarnat's for some time. Specifically for this indication, I think it's really interesting. We talk about this all the time in our state collaborative for referral centers—and many of our listeners are at referral centers. The difference between phone and telemedicine in this situation is that the babies change over time, right? So within the hour or two or four hours that it takes to get a baby, they may look different than when the referring professional sent them. Being able to evaluate that infant by video at that time is very valuable. Because what if you get a baby who now looks pretty good but didn’t before? Or one who looked pretty good and now doesn’t? How does that change what you might do for that infant?

Ben Courchia MD For sure. And like I was saying earlier, for me, I know that the tech is there. You can get very high-resolution video. It’s not going to be like FaceTiming your grandmother across the globe. Sometimes it feels like, “My God, am I really going to have to make a medical decision based on a very pixelated, shoddy type of image that's shaking?” No. The devices are there. They’re stable. And I think it’s important for us to start learning about them because tomorrow, your hospital might say, “We’re going to provide that service to our outlying neighboring hospitals.” So I think it’s going to be an interesting topic and definitely an interesting study, to say the least.

Now I’m going to do an important paper. This was published in the Journal of Pediatrics. It’s called "Norepinephrine versus Dopamine for Septic Shock in Neonates: A Randomized Control Trial." I feel like we’re burying the lead. This is an important paper, and we’re reviewing it at the end of the episode! But it will be available as a Journal Club Short, so that’s okay.

The question is always the same: Which vasopressor do we use as a first-line agent for septic shock in neonates? Sepsis is a major cause of NICU admissions. We’ve already talked about sepsis a lot today just in this Journal Club. It’s a significant contributor to neonatal morbidity and mortality, especially in low- and middle-income countries. Septic shock, in particular, has a high case fatality rate of about 40 to 80%. Management involves timely initiation of fluids and vasoactive support, along with other supportive measures and antibiotics. However, there’s a lack of well-defined management protocols for fluid-refractory septic shock in neonates. Possible reasons include unclear definitions of hypotension, complex pathophysiology, and a lack of robust evidence. Recently-published pediatric and adult sepsis guidelines prefer norepinephrine over dopamine as the first-line vasoactive agent for septic shock, based on adult evidence showing reduced all-cause mortality and better cardiovascular profiles. Several pharmacological factors support using norepinephrine in neonates:

1. The pathophysiology of septic shock involves low systemic vascular resistance, which is addressed effectively by a potent alpha-1 agonist like norepi.

2. Dopamine acts by releasing norepinephrine from sympathetic nerve endings, which may be a limiting factor in sick neonates with limited reserves.

3. Dopamine’s metabolism and clearance depend on postmenstrual age and may lead to drug accumulation in unstable preterm neonates.

4. Norepinephrine may reduce pulmonary vascular resistance and improve pulmonary hypertension—common in neonatal sepsis.

So, the authors conducted this double-blind, parallel-group randomized controlled trial at a tertiary center in northern India over 18 months. They enrolled term and preterm neonates with fluid-refractory septic shock. Each neonate was enrolled only once. Term neonates were eligible until day 28 of life, and preterm neonates until 40 weeks postmenstrual age.

Shock was defined per ACCM guidelines as hypotension (SBP, DBP, or MAP below the 3rd percentile) plus two of the following:

• urine output <0.5 mL/kg/hr over 12 hours

• cap refill ≥4 seconds

• lactate >5

• base excess >5

• SVC flow <40 mL/kg/min

Hypotension was measured via invasive BP monitoring. Fluid refractoriness was defined as no reversal after ≥1 bolus of 10 mL/kg over 10 minutes. Sepsis was defined via positive blood culture, positive screen, or compatible clinical course (plus ≥5 days of antibiotics). Exclusions were congenital anomalies, HIE, prior inotrope use, post-surgical shock, and failure to obtain consent.

Primary outcome was incidence of shock reversal at 30 minutes after starting vasopressors (shock reversal was defined as using American College of Critical Care Medicine guidelines as the correction of hypotension along with any two clinical end points, such as cap refill of two seconds or less, normal pulses with no difference in quality between peripheral and central pulses, and warm extremities). Secondary outcomes included changes in cerebral oxygenation, escalation of vasoactives, steroid use, hyperglycemia (≥180), arrhythmia, respiratory support need/duration, and acid-base changes.

Babies were randomized to norepi or dopamine in 1:1 ratio, stratified to <35 weeks or 35 weeks or greater. All study participants except infant's nurse were blinded. Starting dose of norepi was 0.2 mcg/kg/min and dopamine was 10 mcg/kg/min. Shock reversal was assessed at 15 and 30 minutes. Non-responders received additional agents as per clinical team; crossover was only allowed for refractory cases.

Let's get into the results. 80 neonates were randomized (41 norepi, 39 dopamine). Most had late-onset sepsis (60% in norepi group, 64% in dopamine group). Median age of enrollment was 4 days. About 49% had culture-positive sepsis, so a lot had culture-negative sepsis. Most common isolate was gram-negative organism. Mean gestational age was 33.2 weeks in both groups.

In terms of the primary outcome, the proportion of infants who achieved shock reversal at 30 min were similar in both groups: norepi 31.7%, dopamine 46% (p=0.19, not statistically significant). Requirement of additional vasoactives/steroids were similar. There was no difference in the proportion of neonates attaining shock reversal at any time: norepi 48.8%, dopamine 53.9%. Time to shock reversal was also similar: 1 hour for both. Rates of hyperglycemia was similar. However, tachycardia was significantly higher in dopamine group (15.4% vs 2.4% in the norepi group). There was no difference in IVH grade 3-4, ROP, NEC, or all-cause mortality. Cerebral oxygenation was similar at 30 and 60 minutes in both groups, however at 24 hours, it was lower in the dopamine group. Dopamine group also had significantly lower pH at 24 hours (7.19 vs 7.27; p=0.04). Lactates were similar. Subgroup analysis showed no differences based on GA (<35 or ≥35 weeks), although dopamine did have higher rates of hyperglycemia in the more preterm group.

In conclusion, norepinephrine and dopamine are similarly effective, but dopamine has more side effects: tachycardia, acidosis, lower cerebral oxygenation. Larger studies needed to assess long-term outcomes. I think this adds to the growing data that norepinephrine has a place in our NICU vasopressor toolkit. We probably don’t use it as often as we should. The biggest problem is that it's hard to distinguish between types of shock in neonates. Unlike adults, where you can identify septic vs cardiogenic vs hypovolemic shock more clearly, our babies often have overlapping pathology. The reflex of going to dopamine first, it may be because I don't know what's going on. If you have a baby with a bad IVH, is that hypovolemic shock? Is that baby losing enough blood to actually constitute hypovolemia? Or did the IVH start from septic shock, and that's what I'm dealing with? If you’re fairly certain you're dealing with fluid-refractory septic shock, consider norepi. You’d be supported by strong data.

Daphna Yasova Barbeau MD I had two papers I wanted to highlight, but we are way over time already. What do you think?

Ben Courchia MD Are you going to save them for next time? Or do you want to do a quick one?

Daphna Yasova Barbeau MD How about I just do one? It's a meta-analysis, so I’ll just review the findings in the abstract so people can take a look. It’s titled "All-Cause Mortality and Infection-Related Outcomes of Hospital-Initiated Kangaroo Care Versus Conventional Neonatal Care for Low Birth Weight Infants: A Systematic Review and Meta-Analysis." Lead author Chiara Manzati, senior author Julia Bialicki, published in Lancet Child & Adolescent Health. They evaluated how kangaroo care impacts infection outcomes for low-birth-weight infants. Most guidelines recommend at least one hour daily of uninterrupted kangaroo care, and the WHO actually recommends 8–24 hours/day. In the U.S., we often fall short of that. They included 29 studies, mostly from low- and middle-income countries. Two from the U.S. and one from the UK were included. Some studies were older (1–2 decades), but most were moderate-to-high quality.

In terms of findings, hospital-initiated kangaroo care reduced all-cause mortality (OR 0.77, high-quality evidence), reduced sepsis (OR 0.55, moderate quality), reduced invasive infection (OR 0.49, moderate quality), reduced sepsis-related or invasive infection-related mortality (OR 0.63, high quality), reduced hypothermia (OR 0.28, moderate quality), and reduced apnea (OR 0.46, moderate quality).

Heterogeneity was largely due to variations in kangaroo care offered in the control arms. So even though people argue this mostly applies to low-income countries, I think the same benefits could apply here if we were as committed to prolonged kangaroo care.

Ben Courchia MD Yeah, that’s a great point. The intervention works. We're just not implementing it as well as we could. And resources aren't the issue—it's not like we’re missing some expensive machine.

Daphna Yasova Barbeau MD Exactly. The limiting resource is often the parent. Our system doesn't support parents being at the bedside as much as they need to be. But if they’re there, they should be given the option to do kangaroo care.

Ben Courchia MD Yeah, and the infection data is very compelling. Anyway, thank you, Daphna, for running through all of this, and for another episode of Journal Club. We’ll see you later this week with more Incubator content. And again, feel free to leave a review or a five-star rating on Apple Podcasts—it helps us a lot. Have a great day!